COSC 4368 --- Fundamentals of Artificial Intelligence Spring 2023
( Dr. Eick )
last updated: January 15, 2024
Purpose of this Website
This website intends to satisfy the information requirements of
two independent groups:
- Students who take the undergraduate AI course
- People that want to find out what AI is,
what its subfields are,
and how its technologies, techniques, and
methodlogies can be used in industrial, government, and
research-oriented environments.
If you have any comments
concerning this website, send e-mail
to: ceick@aol.com
Basic Course Information
2023 COSC 4368 Syllabus
(this syllabus will be replaced by a new one by January 17, 2024 the latest!)
class meets: MO/WE 2:30-4p
class room: ....
Instructor: Dr.
Christoph F. Eick
office hours (online using 4368-Class in MS Team): MO 4-5p WE 9-10a
TA: Md. Mahin
TA office hour: MO 1-2p and TU 11a-noon
TA office: online
TA Email: mdmahin3@gmail.com
TA: Raunak Sarbajna
TA office hour: MO noon-1p WE 11a-noon
TA office: online
TA Email:
cancelled classes: none at the moment
makeup classes: none at the moment
lectures taught by others: none yet
Topics Covered in COSC 4368
The course will give an introduction to AI and it will cover Problem Solving (covering
chapter 3, 4 in part, 5, and 6 in part,
centering on uninformed and informed search, adversarial search and games, A*, alpha-beta search, and
constraint satisfaction problems), Learning (covering learning from examples (chapter 18 in part),
deep learning (extra material) and a lot reinforcement
learning (chapter 21, chapter17 in part;)), Reasoning and Learning in Uncertain
Environments (covers chapters 13, 14, 15 in part, and 20 in part, centering on basics in probabilistic reasoning,
naive Bayesian approaches, belief networks and maybe Hidden Markov
Models (HMM)).
Moreover, the course will cover Evolutionary Computing, Game Theory,
Ethics for AI, Deep Learning centering on autoencoders, language models, and deep reinforcement learning relying on other teaching material,
unless the new 4th edition of our textbook now includes coverage of those topics.
Course Materials
Recommended Text:
- S. Russell and P. Norvig, Artificial Intelligence, A
Modern Approach, Fourth Edition,
- Prentice Hall/Allyn&Bacon, December 2020,
-
Link to Textbook Homepage.
Course Elements
There will be a midterm and a final exam in Spring 2024. This semester we will have 3 problem sets which contain tasks which require
programming, and tasks which use AI tools, and an essay writing task. There will be six tasks
in the three problem sets! There will be a 7-week group project which will start approx. February 20, 2024. Finally, each student will be
involved in a single group homework credit (GHC) task (which are also group tasks), whose
solution needs to be presented during the COSC 4368 lecture. Each group will solve a different "kind of homework" problem!
News COSC 4368 Spring 2023
- Currently, you mostly see the course website of the Spring 2023 teaching of this course; this website will
be updated incrementally as we move along with teaching the course in Spring 2024. The course topics covered in
Spring 2024 will not be significantly different from those covered in Spring 2023, except there will be
more coverage of "constructive AI" in Spring 2024!
- The course final exam has been scheduled for Friday, May 5, 2p in SW 101. It will take close to 2 hours
and counts 26% towards your course grade. A detailed review list
has been posted below!
- The course will also use MS Teams as a repository and chat platform. A MS Team for this course should
be available by January 24 the latest. Moreover, if you want to contact Raunak, Mahin or Dr. Eick about something important, please use regular
e-mail and do not use MS Teams!
Important 2024 Dates COSC 4368
We., January 17, 2:30p: First Course Lecture
We., March 8, 2:30p: Midterm Exam (Review List;
March 6, 2023 Review)
March 12+14: Spring Break: no lecture
Mo., March 19: 20-30 minute Lab taught by Steve in preparation of Task3
We., March 21: Lecture on Neural Networks and Autoencoder and Lab in prepation of Task4 taught by Md. Mahin.
Mo., April 30, 2:30p: Last lecture
Fr., May 4,2p(???): Final Exam (Final 2023 Review List (updated on May 1),
May 1, 2023 Review for Final Exam)
Tentative Course Organization
1. Introduction to AI
2. Search
3. Evolutionary Computing
4. Game Theory (very short)
5. Reinforcement Learning
6. Supervised Learning, centering on Basics, Support Vector Machines and Neural Networks
7. Deep Learning (will cover autoencoders, diffusion models, and briefly transformer&language models&GPT Variants and deep reinforcement learning in 2024)
8. AI Politics, Societal and Ethical Aspects of AI
9. Reasoning in Uncertain Environments
10. Planning (only if enough time; not covered in 2023)
2023 Problem Sets and Group Project
Problem Set1 (two individual tasks
centering on search; updated on Feb. 2)
Problem Set2 (two individual tasks
centering on supervised learning and generators/autoencoders; Steve's 2023 Task3 Lecture,
Task3 Jupyter
Notebook, Task4 Jupyter
Notebook; you find the March 22 Autoencoder lecture in the deep learning slides below)
Problem Set3 (Task5
Grading Rubric; Task6: take a
look at Khadija's How to create and use BBNs in Netica video))
2023 Group Project (February 24-April 23, 2023): Learning Paths in a 2-Agent 3D
Transportation World using Reinforcement Learning
(2023 PD World, 2023 Teams)
Tentative Weights in 2023 (subject to change): Problem Set Tasks: 30%, Group Project:17%, Midterm Exam: 21%,
Final Exam: 26%, GHC: 3%, Attendance: 3%.
2023 Group Homework Credit Tasks
2023 Groups
Tasks
Group A and Group B Tasks (both groups will present on We., Feb. 8;
Group B will present a revision of their solution on We., Feb. 15)
Group C Task (will present on Mo., Feb. 13)
Group D Task (will present on We., Feb. 22)
Group E Task (will present on We., March 1)
Group F Task (will present on Mo., March 6 (and maybe March 20))
Group G and H Task (will present on Mo., March 27)
Group I Task (will present on Mo., April 3)
Group J will make a 10-13 minute presentation "Will China be the Number 1 in AI", followed by a discussion,
on We., April 12!
Group N will discuss ChatGPT Mo., April 17
Group K will give a presentation on Robot Soccer on Mo., April 24
Group O will give a presentation on 'AI and Fake News' on We., April 26
Group L Task (will present on Mo., May 1)
Group M will give a presentation about the European AI Ethics Guidelines on May 1
Group Homework Credit (GHC) presentations should take about 12 minutes
and should never be longer than 15 minutes. The presentation
will be streamed in MS Team 4368-Class. That is, you will join the lecture's MS Teams meeting with your laptop, share
your screen and then make your presentation, switching presentators during your presentation. It is okay, if
some of your team members present parts of your presentation remotely. Finally, upload your
presentation slides in the file section of the 4368-Class channel, dedicated to Group Homework Credit.
COSC 4368 Lecture Transparencies
- 2023 Introduction to AI and Course Information
COSC 4368 (will be used for first two lectures in 2023); see also Dr. Eick's 2019 AI Talk"!
- 2023 Search Transparencies:
- Search1 (Classification of Search Problems, Terminology, and Overview
),
Search2 (Problem Solving Agents),
Search3 (Heuristic Search and Exploration),
Search4 (Randomized Hill Climbing and Backtracking; not covered in textbook),
Backtracking Wiki (to be discussed in March 7 review for midterm exam),
Search5: Games (credit for
almost all slides goes to ai.berkely.edu, reduced coverage in 2022),
Search5a (Brief Discussion of Bridge and
Man vs. Machine Game Contests),
Search6: Constraints Satisfaction Problems (credit for
some slides goes to ai.berkeley.edu),
Search6a (Dhar & Quale's paper on Dependency Directed
Backtracking (DDBT); not covered in 2022 but might be useful for Task3 of Problem Set1),
Search7: More on Expansion Search (only material which
centers on greedy search and A* will be covered in 2022),
Search8 (Kamil on Backtracking; not covered),
Suggestions for Solving the Rook+King vs. King Endgame (WRKBK) Problem
(not discussed on 2022).
- 2023 Teaching Material on Evolutionary
Computing (EC): EC1: Introduction
to Evolutionary Computing (by Eiben and Smith covering Chapter 3 of their book)
and EC2:Example: Using EC to Solve Travelling
Salesman Problems, Eiben-Smith Introduction to EA (they
call 'EC': 'EA'!), April 6 EA-paper Walkthrough Notes.
- 2023 Game Theory Slides: G1: Introduction to
Gametheory (USC Economics slide show
by Shivendra Awasthi (???), will be used in the lecture) and G2:
Mo Tanweer Mohammed's Introduction to Game
Theory (not used in the
COSC 4368 lecture).
- 2023 Machine Learning Coverage:
- A Gentle Introduction to
Machine Learning
- Reinforcement Learning: RL1 (Introduction to Reinforcment Learning),
Deep Reinforcement Learning: Neural Networks for Learning Control Laws
(by Steve Brunton; will watch the first six minutes of this video which introduces deep reinforcement learning, and
resume watching the video at 13:49 which discusses Alpha-Go and Other Applications of Reinforcement Learning),
Reinforcement Learning from Human Feedback: From Zero to ChatGPT (by
HugginFace; will just watch 5:50-11:00 of this video which discusses how reinforcement learning is used
to train language models, such as GPT),
2019 Soccer RoboCup, Robo Cup,
RL2 (Using Reinforcement
Learning for Robot Soccer; not covered in 2023), RL3
(Kaelbling's RL Survey Article: particularly, read sections 1, 2, 3, 4.1, 4.2, 8.1 and 9);
Steve Brunton's
Introductory Video to Reinforcement (might be used in 2024).
- Introduction to Supervised Learning (also
called "Learning from Examples")
- Neural Networks:
NN1
(3blue1brown: What is a
Neural Network? (suggest you watch this video, if you did not have any exposure to NN before)),
Neural Networks (Dr. Eick's NN slides),
NN3 (Russel's Introduction to Neural Networks,
not covered in the lecture, but you might take a look at it).
- Support Vector Machines (Review
of the SVM lecture; added on March 24, 2020).
- 2023 Deep Learning Coverage: AutoEncoders (Mahin's March 22, 2023 NN&AutoEncoder Lecture), Transformers (Transformer
Basics;
Steve's April 10, 2023 Lecture on Embeddings and Language Models,
Mahin's April 10, 2023 GPT Overview,
Transformer
Lab (not relevant for 2023 Problem Set Tasks),
Convolutional Neural Networks(CNN; not covered in 2023;
CNN Article, Second CNN Article,
CNN Video for Beginners)
- 2019 Logical Reasoning Transparencies (likely will not be covered 2021, due to the cancellation of 2 lectures and due to
the addition to 1.5 "new" deep learning lectures covering
CNNs, and likely GANs and maybe recurrent neural networks, and that there will be more coverage of Societal and Ethical Aspects of AI in 2021:
- 2023 Reasoning in Uncertain Environments
Transparencies
2023 Societal and Ethical Issues of AI and AI Politics
2020 Planning Slides: Introduction to Planning (based on a lecture by Jim Blythe,
Jose Luis Ambite,Yolanda Gil; not covered in 2023, as we additionaly covered language models in 2023 and added more
depth to the discussion of reinforcement learning and societal aspects of AI)
2017 Last Words
2006/2009 Soft Computing Transparencies
A quick look to Knowledge-based Systems
Foundations of AI (quite short; to
be discussed in the last class of the semester)
2009: Topics covered and
not covered in COSC 6368.
Dec. 7, 2004 Review for the final exam;
2023 Polls
Poll1 (January 30, 2023)
Poll2 (April 26, 2023)
2022 COSC 4368 Poll Results
Poll1
Results (conducted on April 13, 2022)
Old 2023 News Items
- A MS Team called COSC4368 will be associated with the teaching of COSC 4368. Please,
join the team using the passcode jtnu9s7 immediately!
Reinforcement Learning Videos
Please view the following 3 videos:
Siraj Raval: How to use Q Learning in Video Games Easily (7 minutes, will show the first 3:30 on February 20, 2019)
Richard Sutton: Deconstructing Reinforcement Learning (about 50 minutes)
Eric Guimarães:Demo Q-Learning in a GridWorld(2 minutes)
2019 4368 Review Solution Sketches
Solution Sketches April 8, 2019 Review for Midterm2 Exam
Prerequisites
COSC 2320 or COSC 2430.
Other Matial Related to COSC 4368
Some Summaries of
the COSC 4368 Questionnaire Responses from January 23, 2019
Undergraduate Research in Dr. Eick's Research Group
2022 UH-DAIS Research Overview
2023 Lecture Attendance
Attendance counts 3% towards your overall
course grade for
the Spring 2023 teaching of the course. You have to attend at
least 4 F2F lectures in January+February 2023, at least 3 F2F lectures in March and at least 4 F2F
lectures in April/May 2023; otherwise a penalty will be assessed. As MS Teams online attendance records
mostly have been lost in the first 2 months of the semester and
for other reasons, the semester's attendance score will be computed based on F2F attendance as follows:
If you attended the minimum number of F2F lectures in the 3 observation
periods your attendance number grade will be 95; if you satisified the minimum requirement in 2 periods, your attendance number
grade will
be 91, if you satisfied the attendance requirement only in one period you number grade will be 79; if you satisfied the minimum attendance
requirement not in a single period your attendance number grade will be 67. Basically, the attendance
score were computed assuming that you attended all lectures online that you missed F2F! This explains why the
attendance number grades are very high this semester.
2023 Policies Concerning Late Submissions
Submissions up to 24 hours late receive a 10% penalty; submissions 24 hours and 1 minute to 48 hours late receive a 25% penalty,
and submissions received more than 48 hours late will not be graded.
2002 Exam Solution Sketches
March 9, 2022 Midterm Exam A Solution Sketches
March 9, 2022 Midterm Exam B Solution Sketches
Grading
Translation number to letter grades:
A:100-92 A-:92-88 B+:88-84 B:84-80 B-:80-76 C+:76-71
C: 71-66 C-:66-62 D+:62-58 D:58-54 D-:54-50 F: 50-0
Only machine written solutions to homeworks and project reports
are accepted. Be aware of the fact that our
only source of information is what you have turned in. If we are not capable to understand your
approach or solution, you will receive a low score.
Moreover, students should not throw away returned assignments or tests.
Students may discuss course material and homeworks, but must take special
care to discern the difference between collaborating in order to increase
understanding of course materials and collaborating on the homework /
course project
itself. We encourage students to help each other understand course
material to clarify the meaning of homework problems or to discuss
problem-solving strategies, but it is not permissible for one
student to help or be helped by another student in working through
homework problems and in the course project. If, in discussing course materials and problems,
students believe that their like-mindedness from such discussions could be
construed as collaboration on their assignments, students must cite each
other, briefly explaining the extent of their collaboration. Any
assistance that is not given proper citation may be considered a violation
of the Honor Code, and might result in grade reduction, obtaining a grade of F
in the course, and in further prosecution.
2020 Reviews and Exams
Midterm1 Exam (Mo., March 2, 2020 Review List,
February
26, 2020 Review for Midterm1 Exam)
Midterm2 Exam (Mo., April 13, 2020 Review List,
April 8, 2020 Review
for Midterm2 Exam (only 40% of the review questions will be discussed on April 8!)).
Final Exam (Mo., May 4, 2p, Review List 2020 Final Exam,
April 27,
2020 Review for Final Exam, Solution Sketches May 6, 2019 Final Exam)
2021 Final Exam and Review for it
Final Exam (We. May 12, 2022 2p, First Draft of Review List 2021 Final Exam (will
be finalized by May 6 the latest,
May 3
2021 Review for Final Exam, Solution Sketches May 6, 2019 Final Exam)
2022 Problem Sets and Group Project
Problem Set1 (contains two individual tasks centering on search)
Learning Paths in a 2-Agent Transportation World using Reinforcement Learning (group project
centering on reinforcement learning; duration:
February 28-April 26, 2022; 2022 PD World;
2022 Groups) updated on April 6, 2022)
Problem Set2 (updated on March 28, 5p;
two individual tasks: Using Neural Network and SVM for a Classification Task and using Transformers for a
sentiment analysis problem; Steve's Transformer Lecture, Transformer
Lab)
Problem Set3 (preliminary first draft; Task5 Grading Rubric)
Tentative Weights for the different parts of the course in 2022: Problem Sets: 30% , Group Project: 17%, Midterm Exam:21%,
Final Exam: 25%, Attendance: 4%, Group Homework Credit: 3%.
2021 Problem Sets and Group Project
Problem Set1 (contains individual tasks centering on search, typo in equation C8
has been correct on Feb. 19, 2021 at 1p)
Problem Set2 (contains individual tasks centering on neural networks, support vector
machines and artificial generative networks)
Problem Set3 centering on decision making in uncertain
enviroments and ethical and societal aspects of AI (containing "individual" tasks:
no collabortion with other students allowed; Essay Evaluation Criteria for Task 6; take a
look at Khadija's How to create and use BBNs in Netica video)
Learning Paths in a Transportation World (group project centering on reinforcement learning; duration:
February 25-April 11, 2021; 2021 PD World)
The "tentative" weights of the problem sets tasks are as follows: Task1: 12, Task2: 33, Task3: 35, Task 4: 25, Task 5: 28(??),
Task 6: 16 and Task 7: 18.
Older News Items
The 2023 attendance policy has been posted below; you are expected to attend 44% of the lectures
F2F and the remaining lectures you can attend online. Attendance will count 4% towards your course grade.
Moreover, if you miss a small number of lectures that will not have much impact on
your attendance grade.
Miscellaneous
Finally, Congratulations go to all 4368 students who graduated in Spring 2020
semester!!! This slide is part of the "must see"
Spring 2020 NSM Graduation Celebration video.
